Dr Thomas Robinson

This course is available on the MPhil/PhD in Computational Social Science and MPhil/PhD in International Relations. This course is freely available as an outside option to students on other programmes where regulations permit. It does not require permission.

This course is available to research students only. This course is not controlled access. If you register for a place and meet the prerequisites, if any, you are likely to be given a place.

Applied Regression Analysis (MY452) or equivalent is required. Students should have a good grasp of at least one programming language. If this programming language is not R, students should take the Digital Skills Lab course in R before the start of term.

Machine learning uses algorithms to find patterns in datasets and make predictions from them. This course focuses on how these methods can augment, extend, or complement scientific research pipelines—treating prediction not as an end in itself but as a tool for inference and measurement. Lectures develop the fundamental concepts that run across machine learning-- including generalisation, the bias-variance tradeoff, and regularisation—illustrated through strategies including regularised regression (e.g. LASSO), tree-based methods, distance-based algorithms,neural networks, and unsupervised learning. Students will engage critically with the benefits and costs of these methods, including questions of algorithmic bias, fairness, and privacy, drawing on prominent examples from social science research.  In seminars, students will build and apply algorithms to data and validate and evaluate models, working directly with social data in Python or R.

20 hours of lectures and 10 hours of seminars in the Winter Term.

This course has a reading week in Week 6 of Winter Term.

This course is delivered through a combination of classes and lectures totalling a minimum of 20 hours across Winter Term. 

Students will be expected to submit 1 problem set in WT.

• Géron, A. (2017). Hands-on Machine Learning with Scikit-Learn and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems. O'Reilly Media, Inc.
• Müller, A. C., & Guido, S. (2016). Introduction to Machine Learning with Python: A Guide for Data Scientists. O'Reilly Media, Inc.
• Conway, D., & White, J. (2012). Machine Learning for Hackers. O'Reilly Media, Inc.
• James, G., Witten, D., Hastie, T., & Tibshirani, R. (2013). An Introduction to Statistical Learning (Vol. 112). New York: Springer.
• Cantú, F., & Saiegh, S. M. (2011). Fraudulent democracy? An analysis of Argentina's Infamous Decade using supervised machine learning. Political Analysis, 19(4), 409-433.
• Davidson, T., Warmsley, D., Macy, M., & Weber, I. (2017). Automated hate speech detection and the problem of offensive language. Proceedings of the Eleventh International AAAI Conference on Web and Social Media (ICWSM 2017), 512-515.
• D'Orazio, V., Landis, S. T., Palmer, G., & Schrodt, P. (2014). Separating the wheat from the chaff: Applications of automated document classification using support vector machines. Political Analysis, 22(2), 224-242.
• Jones, Z. M., & Lupu, Y. (2018). Is There More Violence in the Middle?. American Journal of Political Science, 62(3), 652-667.
• Kosinski, M., Stillwell, D., & Graepel, T. (2013). Private traits and attributes are predictable from digital records of human behavior. Proceedings of the National Academy of Sciences, 201218772.
• Wang, Y., & Kosinski, M. (2018). Deep neural networks are more accurate than humans at detecting sexual orientation from facial images. Journal of Personality and Social Psychology, 114(2), 246-257.

Exam (80%), duration: 120 Minutes in the Spring exam period.

Problem sets (20%).

One summative take-home assessment in WT (20%) and one exam in ST (80%).